Left atrial remodeling: assessment of size and shape to detect vulnerability to atrial fibrillation

The ellipse formula often underestimates left atrial (LA) dilation. Complete echocardiographic analysis of LA shape in relation to electrical remodeling has not been performed.

To analyze the relation between LA shape/surface and vulnerability to atrial fibrillation (AF).

We studied 112 patients aged 43 +/- 16 years, and referred for electrophysiological study. LA surface (LAS) was measured at end-systole (maximal). Trapezoidal LA shape was defined if the transverse dimension was less than the basal dimension. Decremental index (DI) was calculated as the maximal percentage prolongation of interatrial conduction time during atrial extrastimulation. The LA was considered vulnerable if AF was inducible. DI > 50%, repetitive atrial activity, and fragmented electrograms defined susceptibility to vulnerability.

LAS ranged between 10.5 and 36.6 cm(2); 77 patients had a trapezoidal LA. By simple regression analysis LAS correlated with DI (r(2) = 0.38, P = 0.0001). LAS predicted susceptibility to vulnerability better than vulnerability to AF (area under the ROC curve: 0.93 vs 0.81). The best cut-off value for LAS as predictor of susceptibility to vulnerability was 19.5 cm(2) (sensitivity: 89%; specificity: 90.5%; positive predictive value: 93.4%; negative predictive value: 84.4%). Using LAS > 25 cm(2) as a cut-off value, LA vulnerability to AF was detected with a sensitivity of 56.2% and a specificity of 95% (positive predictive value: 81.8%; negative predictive value: 83.3%). LA shape was trapezoidal in 72% patients with LAS > 25 cm(2) and in 30% patients with LAS < 19.5 cm(2) (P < 0.0001).

LA dilation and electrical remodeling are related. Progressive LA dilation is accompanied by shape remodeling. Appropriate characterization of LA remodeling should therefore include LAS measurement and LA shape assessment.